Hi Juha,

>>I think we can add some words about the case if keep=yes is not
>>returned. But, if we for example talk about the UA still sending STUN
>>requests I also think we need to say something about the potential
>>issues also.
>
>not just "some words", but a clear recommendation (at least for the CRLF
>case).
 
Yes.
 
>i don't know what the issues with stun might be (possible proxy crash or 
>blacklisting do not qualify).

I agree that proxy crash doesn't qualify (as you said earlier, such proxies 
should not be deployed in the first place), but I DO think the blacklisting is 
a valid real-life issue.

There is also another thing: I guess one reason the edge proxy doesn't return 
keep=yes could be because it detects (using whatever mechanism) that the UA is 
NOT behind a NAT. In that case the UA may end up sending unecessary 
keep-alives, in case the UA itself doesn't know whether it's behind a NAT or 
not.

 >>Also, if the UA for example receives a very short registration
 >>refresh timer (since the edge proxy did note support "keep"), there
 >>probably is no reason to send keep-alives in addition to the
 >>registration refreshes.
>
>you can write that if registration interval would be as short as
>default keepalive interval then it is not recommended to send
>keepalives.
 
Yes.
 
Regards,
 
Christer
_______________________________________________
Sip mailing list  https://www.ietf.org/mailman/listinfo/sip
This list is for NEW development of the core SIP Protocol
Use [EMAIL PROTECTED] for questions on current sip
Use [EMAIL PROTECTED] for new developments on the application of sip

Reply via email to